© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 66982
    Prev
    Next
    INDEX
    Explanations

    numbers and math symbols

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron primarily detects the literal token “false.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     on
    -1.43
     at
    -1.41
     in
    -1.40
     just
    -1.38
     In
    -1.32
     with
    -1.31
     provide
    -1.24
     When
    -1.19
     As
    -1.19
     six
    -1.18
    POSITIVE LOGITS
     certamente
    1.30
     wacky
    1.28
     insanely
    1.26
     trône
    1.26
     niedersachsen
    1.24
     luscious
    1.23
    ほとん
    1.23
     récol
    1.20
     AMAZING
    1.19
     ridiculously
    1.18
    Activations Density 0.031%

    No Known Activations